Re: [life] reality of perception

["John Kineman" <***>]

John C. and Georg and list, (also cross-posting)

Interesting discussion in which I appreciate discovering new references 
and perspectives. It seems that each wrinkle on these related questions 
is accompanied by its own characteristic literature, and these may not 
overlap or cross-fertilize very well. I also get the impression that 
many treatments over-specify the problem, trying to reason beyond 
experience. My approach, since about 1985, has been to try to synthesize 
a consistent view of science, and then to interpret that as itself a 
natural process. I'll summarize and show some similarities with the 
Rosen view.

As Georg says, this does recapitulate the earlier discussion regarding 
Berkeley, "real" events, and perception, of which most of us are 
familiar. I had started with a quote from Descartes that all we can know 
of for certain is our own experience, all else being subject to doubt. 
Then there was discussion of how well we can know our own experience.

I think Descartes probably knew we could doubt everything if we chose 
to, including our own experience. However, I think he was proposing a 
fundamental value judgement to establish a form of realism. The point is 
not so much what we can know, but what we cannot know. I interpret it as 
saying we can have no more fundamental or authoritative conformation for 
reality than our own experience, for (a) nothing else is available to us 
and (b) the pupose of knowing is already referenced to our experience. 
Thus we accept our percepts, but still question their interpretation. In 
so doing, a scientific method can be established to look, not for truth, 
but for generality, consistency, parsimony, formalization, necessity, 
and usefulness. These six criteria give a means for accepting 
assumptions about nature or changing them. Then, with foundations 
meeting those criteria, we can specify processes (theories) and test the 
veracity of the specification by applying the same six criteria, plus 
falsifiability within the context of the assumptions (worldview) 
established prior (HD methods based on Bacon, Popper, etc.). Proceeding 
thus, we can fill out the range of applicability of a theory to 
perceived phenomena (which has worked best in physical science). Not 
perhaps forseen by Descartes and the positivists, but nevertheless 
revealed in practice, any such program will necessarily discover the 
boundaries of its application to natural (perceived) phenomena - the 
assumptions tend to break down outside of some domain of relevance. That 
discovery will appear in the form of a logical contradiction within a 
theory itself, simply because it is a logical construction on a limited 
premise. When that occurs, and can be demonstrated, it is then possible 
to justify a change in assumptions (Kuhn's revolutions). Hopefully we 
can find one that will eliminate the contradiction,  if such assumptions 
can be found (e.g., Einstein). If not, we end up with different, 
well-founded views and their related theories that each tell us a 
different aspect of nature (e.g., instrumentalism as described by Peters 
in a critique of biology, others).

In this process is embedded a form of what I think of as "relative 
realism:" Reality is conceived in relation to the assumptions and theory 
elements established. Change the world view (assumptions) and one's 
concept of reality changes. Testing of theories is then shifted to a 
different basis. True and false are thus defined by the context of one 
view of reality or another. This also leads to a "provisional 
instrumentalism," the provision being that unless one is indeed looking 
to meet the principles stated above, which itself defines a concept of 
reality (presumably the most basic one possible), then fragmentation can 
go wild and connections between new constructs and old ones will not be 
required, making it impossible to understand one view in terms of 
another (not suggesting they fully correspond, but that one can ask what 
the new information looks like in the more familiar terms).

All this is a recapitulation of a philosophy of science that I have 
cobbled together from various readings and thoughts since 1985, but it 
seems to support our conclusion here, that accepting one's own 
experience (not explanation) of an external world, while the truth of 
that proposition is ultimately undecidable, is nevertheless the only 
option that preserves reason. So, we assume there is a noise in the 
forest with or without us, not because there is any closed logic 
requiring it, but because to not equate reality with repeatable 
experience is to not think about one's experience, which in any case we 
consider important. And to not apply epistemological criteria to our 
explanations of the experience would mean being inefficient, fragmented, 
or fanciful about the explanation.

This much might be accepted at least as one interpretation of science, 
but there is now more that I will get into, because of our question 
about life. Let me try to phrase that new element in related terms, 
referencing Rosen now in a similarly general way.

The above approach, in various forms and subsets, has been applied to 
build science as we know it today. But now we look at systems that, in 
effect "do their own science." This is one of the big insights I gain 
from Rosen's theory, that organisms contain their own models and 
modeling processes. If we think of it that way, we are asking not just 
about epistemology (the art and rules of knowing) but about knowing 
about knowing - epistemology of epistemology. That is obviously a 
compound language element, and it should not be difficult to imagine 
such a compound description indicating a complex phenomena that loops 
onto itself in important ways producing a much less definable result 
than epistemology alone. If we try to study a scientific system itself, 
the rules become difficult to follow, because we have to consider the 
effect of the rules on the same rules, while we apply the very rules we 
are attempting to discover. So it must be a highly iterative process 
that tries to step outside its own rule set, and there is no longer any 
guarantee of universality. Clearly we need the most general rule set 
imaginable to tackle this second-order problem, and accordingly Rosen 
argues for removing all the restrictions previously placed on 
mathematics to produce computability and to assume causality.

Some of the epistemological criteria may be retained, but some require 
modification. Parsimony is best understoood (IMO) as explanatory 
elegance. We can still hope for that and recognize it when it appears, 
if we do not think of it a "simplicity" alone. In Occam's words it is 
not overcomplicating the explanation beyond what is needed to capture 
the phenomenon, which in this case now includes unknown behaviors that 
can nevertheless be expected. So, parsimony doesn't mean eliminating 
uncertainty or only constructing simple answers to complex questions. 
Consistency means we should not contradict prior experience or 
experience from multiple perspectives (explanations can be contradicted, 
but not confirmed experience). Again, one now has to make new allowances 
for the emergence of novel behavior, which should not be taken as 
inconsistencies; but otherwise we can retain the critieria of 
consistency across experience.  Generality means that we think general 
theories are preferred to system-specific ones. The ethic may be 
retained in principle, but is seriously diminished in practice. There 
are now many aspects of knowing that must be case specific, as "knowing 
systems" themselves generate unique system behaviors. We now have to 
consider history itself in explanations. So we give up on finding 
universal laws about such systems as a practical or detailed matter, but 
may still look for ways of knowing about groups of systems in general 
terms. A hierarchy of explanations thus matches a hierarchy of perceived 
systems. Generality is now achieved through reasoning by analogy 
(something Rosen also emphasized the importance of). Formalization - the 
ability to formalize a theory - also comes into question because the 
traditional method is predicated on being able to write down 
descriptions of nature and use those descriptions again. When doing this 
for systems that are writing their own descriptions of self and nature, 
our description of the whole organism would have a hard time capturing 
that compound element. Rosen found that to approach this problem 
requires moving out of the purely quantitative domain and looking at a 
broader range of mathematics that includes category theory. So there are 
some possibilities regarding formalization. Finally, Necessity and 
usefullness are criteria than can obviously be retained and that in fact 
become more important in a complex environment where there are too many 
possible simultaneously valid views and descriptions to be all 
identified. Certainly, constructing them on a priority basis makes sense.

This was an attempt to map the discussion into a view based on 
epistemological criteria and a model of science itself. To me this is 
the most interesting implication of Rosen's work, as it starts with an 
epistemological relationship and then uses that to explain the 
complexity of organisms. In so doing, it reveals implications for how we 
imagine reality  itself - that we must now think of it as fundamentally 
complex, although not fragmented as in modernism and post-modernism. 
Since the theory is based, most fundamentally, on the concept of 
percepts, it is implicitly consistent with Descartes premise.

All the Best,
John Kineman

John Collier wrote:

>Georg wrote:
>
>John wrote:
>
>This leads to an antirealism about at least some things -- our perceptions always go 
>beyond the information in them, at least insomuch as they are classified in any way -- 
>and we don't have any unclassified perceptions, inasmuch as they are meaningful in any 
>sense that allows us to group them together in some way or another. So it seems to me 
>that the usual forms of realism and antirealism are both false. 
>
>Georg:
>
>Do you mean by antirealism to say that things are not? Or not as we perceive them? I 
>think both approaches are wrong. In 'postmodern' thinking people often maintain that 
>they don't perceive what they perceive. I find this disastrous.
>
>John:
>Likewise. Anitrealism is hard to define, since the terms required shift meaning from a 
>realist context to an antirealist one, but the basic idea is that we construct all that 
>we think an perceive, and that truth is a matter of some definable decision procedure 
>(e.g., negotiation, power, verification). I should note that Tarski showed that truth as 
>it is used in logic cannot be defined formally, so the antirealist is typically not 
>talking about truth as it is used in logic. What they are talking about I suppose is 
>negotiation, power, verification or something of the kind, and I wish they were more 
>honest about what they are talking about. For example, some logicians and philosophers, 
>e.g., Michael Dummett, use true in a system, but then let that slide over into truth 
>simplciter. True in a system need not be true at all, though we usually give people the 
>benefit of the doubt on this for purposes of communication(Principle of Charity).
>
>Georg:
>You introduced here the term information. As I understand that is that a tree gives us 
>the 'information' of a tree but perception connects it with many other ideas. Correct?
>
>John:
>Basically.
>
>Georg:
>I would see it the other way round. All the 'ground' of our experience gives rise to the 
>gestalt of the tree. This is about the same discussion I had with John Kineman on the 
>question whether or not there are facts. This is here translated into information.
>
>John:
>I think it is the tree that makes the difference, while the ground remains the same, so 
>I would say that the perception contains information about the tree. Information is a 
>distinction that makes a difference (Donald Mackay). The ground makes no difference in 
>this case, as it remains the same, so it provides on information.
>
>John
>
>
>Professor John Collier  
>Philosophy, University of Natal
>Durban 4041 South Africa
>T: +27 (31) 260 3248 / 260 2292
>F: +27 (31) 260 3031
>email: ***
>
>
>--------------------------------------------------------------------
>Please find our disclaimer at http://www.disclaimer.nu.ac.za
>--------------------------------------------------------------------
><<<<gwavasig>>>>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: ***
>For additional commands, e-mail: ***
>
>  
>